Advanced age is a major risk factor for cognitive decline. Accumulating evidence demonstrates the importance of physical activity for preserving brain health in older adults. However, at present, we know very little about the underlying mechanisms. Arterial stiffening and increases in central pulse pressure are the hallmark of arterial aging which may have significant impact on brain perfusion. The overall objective of this proposal is to test the central hypothesis that exercise training improves brain perfusion by modifying arterial aging and these changes lead to increases in regional brain volume and reductions in white matter lesions, and thus improvement in cognitive function in older adults over 65. The following specific aims will be accomplished: 1) to determine whether increases in arterial stiffness and central pulse pressure with age are associated with increases in cerebrovascular resistance/impedance, reduction in brain perfusion and cerebrovascular dysfunction; 2) to determine whether exercise training improves brain perfusion by modifying arterial aging as indicated by reductions in cerebrovascular resistance/impedance, arterial stiffness, central pulse pressure and/or improvement in vascular endothelial function; 3) to determine whether improvement in brain perfusion with exercise training leads to increases in regional brain volume, reductions in white matter lesions and improvement in cognitive function. To accomplish this goal, a cross-sectional study of the young, middle-aged and elderly individuals will be conducted for aim 1. Furthermore, a longitudinal study of aerobic exercise training and control (flexibility and balance training) will be conducted for one year in the elderly for aim 2 and 3. The state-of-the-art transcranial Doppler and magnetic resonance imaging will be used to measure changes in brain perfusion and structure. A comprehensive battery of neuropsychological testing sensitive to age will be used to measure changes in cognitive function. Advanced statistical methods of multiple variable regression and mediation model analysis will be used to determine the relationship between arterial aging, brain perfusion, structure and cognitive function.